Significance testing of this dataset was performed using linear models and pair-wise comparisons [ 116].
When our model was tested on this dataset, an AUC of 69.51% was obtained (blue curve in Figure 3), suggesting that the model had the potential to achieve the goal of predicting potential microRNA-disease associations for the diseases without any known related microRNAs.
To test whether the results of the analysis are significant, we perform permutation testing of the dataset.
Following plausibility testing of the dataset, the normality distribution was checked by using the Kolmogorov-Smirnov test and was approximately fulfilled by all variables.
To test the classifiability of this dataset, we calculated the log likelihood ratio scores of all the words in it, including the true binding sites, and tried out various threshold values to classify the words.
We have tested the "noisy" variation of this dataset using the HOGHOFMBH model after recalibrating the Platt scaling using four-fold cross-validation.
Testing of this algorithm on a comprehensive dataset showed significant improvement on prediction accuracy upon currently widely-used programs.
Each discretization algorithm performed with pairs of 10 training and 10 testing of each dataset.
We carried out systematic testing of our dataset to find any missing data and extreme values.
However, a reanalysis of this dataset, specifically testing for EGTs, showed no such evidence for red algal contributions to the oomycete genome (Stiller et al. 2009).
(i) We have made a large dataset of 501 peptides with different modifications and tested the performance of PEPstrMOD on this dataset rather than testing on 8 peptides.
